Hao Yan

About Me

I am a Ph.D. candidate in the Department of Computer Science and GMU NLP Group at George Mason University, advised by Dr. Ziyu Yao, where I work on making Large Language Models safe, trustworthy, and usable for everyone. My research covers four connected areas.

1. 🛡️ LLM Safety & Secure Code Generation. I study when and why LLMs write vulnerable code. My work explores how models can be guided to find and fix their own security flaws.
2. 🔬 LLM Interpretability & Explainability. I look inside LLMs to understand how they actually work. My recent work interprets how LLMs generate code and uses these insights to guide them toward safer and more reliable code.
3. 🤖 LLM Reasoning & Agentic AI. I study how well LLMs reason and follow instructions. My work examines whether models can faithfully simulate personas with different skill levels in math reasoning. Beyond reasoning, I have a strong interest in Agentic AI, especially in how multiple agents collaborate and how agents manage their memory over long-horizon tasks.
4. 🤝 Human-AI Interaction & Conversational Code Generation. I design systems that let people create software by talking to LLMs in plain language. I built a conversational assistant that helps non-programmers write correct code, and I taught LLMs to simulate real user feedback so that interactive systems can be trained at scale.

In the summers of 2025 and 2026, I worked at Amazon in Seattle as an Applied Scientist Intern, where I gained hands-on experience applying LLM research to real-world products at industry scale. Prior to my Ph.D., I received my M.S. in Computer Science from George Mason University (2020) and my B.E. in Electronic and Information Engineering from Fujian Normal University (2017).

Outside of research, I enjoy basketball, badminton, and hiking. I am also a foodie who loves exploring new restaurants and cooking at home.

🔍 I am on the job market, seeking full-time positions as an Applied Scientist, Research Scientist, or related roles in industry and academia. Please feel free to reach out via email!

Research Interests

• LLM Safety & Secure Code Generation
• LLM Interpretability & Explainability
• LLM Reasoning & Agentic AI
• Human-AI Interaction & Conversational Code Generation

News

• [Aug. 2026] Our paper “Interpreting and Steering for Safe and Correct Code Generation” is accepted to EMNLP 2026.
• [Aug. 2026] Our paper “Guiding AI to Fix Its Own Flaws: An Empirical Study on LLM-Driven Secure Code Generation” is accepted to ACM Transactions on Software Engineering and Methodology (TOSEM).
• [May 2026] I am returning to Amazon in Seattle this summer as an Applied Scientist Intern.
• [Aug. 2025] Our paper “Can LLMs Simulate Personas with Reversed Performance?” is accepted to the COLM 2025 Workshop on Social Simulation with LLMs.
• [May 2025] I will be interning at Amazon this summer in Seattle as an Applied Scientist Intern.
• [Apr. 2025] Check out our new preprint about LLM persona simulation
• [July 2024] I will present at The EPIC SPICE Intern Program about “LLM agent: build your own personal assistant”
• [May 2024] I received the summer GRA Fellowship from CHAMP GMU
• [Apr. 2024] Check out our new preprint about the HCI for Non-professional programmers in LLM-driven code generation.
• [June 2023] I will serve as a student volunteer at ACL 2023.
• [May 2023] I was granted the “Graduate Student Travel Fund (GSTF)” from GMU.
• [May 2023] Our paper “Learning to Simulate Natural Language Feedback for Interactive Semantic Parsing” is accepted to ACL 2023.

Publications

Experiences

• May. 2026 - Aug. 2026: Applied Scientist Intern, Amazon, Seattle, WA
• May. 2025 - Aug. 2025: Applied Scientist Intern, Amazon, Seattle, WA
• Jan. 2025 - Apr. 2025: Machine Learning Engineer Intern, Salesteq, Remote
• May. 2021 - current: Graduate Research Assistant, George Mason University
• Aug. 2020 - May. 2021: Graduate Teaching Assistant, George Mason University
• Feb. 2020 - Aug. 2021: Front-end Developer, Bright Sky Tech Inc
• Jan. 2020 - Dec. 2020: Graduate Teaching Assistant, George Mason University

Services

Program Committee

• IJCAI'26, IJCAI'25, ECAI'25

Reviewers

• NeurIPS'26, ICLR'26, ICLR'25, ACL ARR'26, ACL ARR'25, IEEE CI-NLPSoMe'25, WLLFM'24, IEEE BigData'24, NeurIPS'23 (Secondary)

Student Volunteer

• The 61st Annual Meeting of the Association for Computational Linguistics (ACL), 2023

Talks

• [July. 2024] Title: LLM agent: build your own personal assistant, Invited talk at The EPIC SPICE Intern Program. Host: Amadra Shehu.

Awards

• Graduate Student Travel Fund, GMU, 2023
• Distinguished GTA Award, Department of CS, GMU, 2022
• Graduate Students Poster Award (Honorable Mention), GMU, 2022
• Outstanding GTA Award, Department of CS, GMU, 2020, 2021

Powered by Jekyll and Minimal Light theme.